Cohere's CEO, Aidan Gomez, raised concerns about the concentration of power in defining AI safety rules. He warned that a few Silicon Valley companies are shaping global AI safety standards without public input, citing past regulatory failures.
Gomez highlighted the risks of allowing a small group of companies to set safety standards, noting that these entities have a history of protecting their interests rather than the public good. He pointed to past regulatory failures in areas like bond ratings and automotive safety as cautionary tales.
The CEO emphasized that while AI needs guardrails, the debate should focus on who sets those rules and whose interests they protect. He argued that the current proposals risk entrenching dominant companies by limiting competition and stifling innovation.
"We’ve been here before," Gomez said. "The stated goal was safety, but the result was a market structure that protected incumbents and limited competition."
The discussion comes amid recent efforts by Silicon Valley incumbents to shape AI regulation, including a roadmap proposed by Anthropic CEO Dario Amodei. Gomez criticized the proposal for lacking public consultation and for allowing a small group of companies to dictate the rules.
Cohere did not say how to address these concerns, and the open question remains about whether the public should have a say in shaping AI safety standards. The CEO called for a more inclusive approach to regulation.
Source: cohere
